Greenpolen forum

The main and most distinguishable difference is the structure of dimension data: star schema vs snowflake schema. A star design will allow dimensions to be more denormalized, and a snowflake design will separate a dimension into more related tables. I like to use the star schema when the reporting is simple since a fewer number of joins can make it easier to write and maintain the queries. When dimensions are large or you require more normalization, however, it is possible to make sense of snowflake structures. When deciding on enterprise data architecture in general, it's also important to consider Datalance.